Bending of structural aluminum standard profiles

[KawaMinaii]

Hello every body.
I want to make a curve beam from a structural aluminum profile. The radius of curvature of the resultant beam should be 2.5 meters and its span will be 8 meters.
If you have any suggestion about the forming processes and the required analysis please guide me through this.

 

[drawoh]

KawaMinaii,

Aluminium extrusions generally are heat treated, and the yield stress and ultimate stresses are not far apart. You cannot bend them very much unless you anneal them.

Read up on your bending theory. When you apply a moment, the material bends around a centroid, into a circle. Your maximum strain is at the fibre the furthest distance from the centroid. This strain is easy to work out.

--
JHG

 

[berkshire]

KawaMinaii (Civil/Environmental) (OP)
When I worked for an architectural sheet metal shop, we also worked with extruded profiles, simple curves within the bending tolerances of the extruded material were done on a "Buffalo 4 roller profile bending roller. Tighter and complex curves were sent to a stretch former with heat treat capability.
B.E.

You are judged not by what you know, but by what you can do.

 

[JStephen]

If you already have the shape in mind, contact somebody that does the rolling and see if they can do it.

It may depend on what equipment they have, and there may be a limit to how tight it can be rolled. It may take experimenting to find out.

I recall several years back, we had some carbon steel channels rolled, and it separated some of the flanges from the webs. With extrusions, I would expect problems if it was too thin or too deep, or an odd shape that doesn't conveniently fit the rollers at hand.

 

[ruhlich]

I work with smaller sizes of aluminum, 99% are 6061-T6. In a design I created, I have to bend tabs on some 1/8" x 2" flat bar. The failure rate was over 50%........so I guess you could say I was exceeding the elastic limit, as was mentioned by someone else. Not knowing any better, I decided to try some heat. I used a propane torch to heat the area to be bent, and viola!!!.....not one fracture in nearly a 100 bends.

If you can spare "trial and error" material, I would suggest you try this as well. How much heat?....No idea, but it doesn't have to change color nor be next to dripping.

As far as the radius goes, I would make a curve template by cutting the curve into a piece of material (aluminum plate or sheet, wood, etc) as large as the CNC Mill will let you (the larger the piece, the more exact your bend comparison). Then as you roll your prototype, you can gauge the amount needed by matching the template curve to the bend.

 

[berkshire]

ruhlich (Mechanical)
when you heat the 6061 material with a torch, you are doing a process called Hot forming, with 6061 material you should not exceed 400 degrees F for more than 5 minutes.
B.E.

You are judged not by what you know, but by what you can do.

 

[Compositepro]

6061-T6 cannot be formed much. Immediately after extruding it is soft in the T-0 temper and can be easily formed. With time it age hardens to T-4 and is still fairly formable. Heating to 350F or so for a few hours hardens it to T6 where it has a much higher yield strength but low strain to failure. Heating T-6 above 400F will soften and weaken it. It is not really annealed, but the grain structure grows large and lumpy. 6061 must be solution annealed at about 1000F where the alloying components dissolve back into solution. Quenching this, cools the aluminum before grains can grow and you have soft, annealed 6061-T0 again. Grains will slowly grow at room temperature. It is the interlocking grains that harden and strengthen 6061.

 

[berkshire]

Following on from Composite Pros remarks the hot forming has the effect of a partial anneal, this can be done between 270 and 400 degrees F.
This has the effect of removing some of the effects of heat treatment and work hardening, but not all, and retaining fair mechanical properties of the metal. The metal will then age back to its original properties.
Excessively High temperatures should be avoided because considerable loss of strength will occur.
B.E.

You are judged not by what you know, but by what you can do.

 

[IFRs]

Please tell us how many you need to make, what shape you are rolling and in what direction.

rulich -
Bending plate is one thing - bend across the grain and use bend radii as recommended by the Aluminum Association, should be no problem.

KawaMinaii:
I roll custom 6061-T6 extrusions all day using a custom roll that contains the flanges. Rolling to a 2.5 meter radius may give you trouble depending on the shape and roll direction. Try it without heating - you have nothing to lose and it may work. If you are rolling an I-beam the easy way, you are essentially rolling two plates the hard way and the web will just follow along. This should work with some trial and error. Look for stress cracks at the outer edges and thickening at the innter edge.

For that small a part and that tight a radius you may want to water jet cut flanges, roll a web and weld it all together.

If all else fails, buy the shape in T0 condition, roll it and then send it out to be heat treated.

 

[tbuelna]

What is the exact profile and section thicknesses of the material you are forming? Is it an I-beam, a channel, a tee, a hat section, an angle? A 2.5M radius does not seem too difficult. But some shapes can be more difficult to roll form without buckling than others. The direction the shape is rolled can also make a big difference. For example, a channel that is rolled around the shear web with the legs facing outward can be rolled to a tighter radius without buckling than if the same channel is rolled with the legs facing inward.

 

[dhengr]

KawaMinaii:
You seem to resist telling us what you are really doing, so we have an idea of the size, actual shape of the alum. shape and the radius of the bends. It’s really tough to have a meaningful discussion on the matter without this info. Why not buy the alum. extrusions in the soft condition, form them, and then temper them?

 

[KawaMinaii]

Thank you all for replying.
Actually I need a few of them, maybe ten or fifteen.
The most difficulty with us is the lack of instrumentation. Maybe I would change the material to steel.
There is a traditional way in which some sectors are removed from the profile in the peripheral direction, then the remaining parts are bent and welded together. Does anyone have a same experience? How could I carry it out?

 

[MikeHalloran]

You still haven't told us the size or nature of the actual profile you wish to bend, so more specific answers are not possible.

In general, cold bending is preferred because it doesn't substantially change the mechanical material properties.

Removing pie-shaped wedges and re-joining after localized bending works fine for for steel, and for wood. For aluminum, not so much, because welding leaves the HAZ quite soft, so the analysis gets complicated because the material properties around the weld are substantially different from the bulk properties, so you are looking at a complicated composite incrementally curved beam.

You might be better off annealing, bending it soft, and heat treating, or just building the curved section up from plate, welding that, and heat treating, or something else entirely, depending on information not revealed so far.





Mike Halloran
Pembroke Pines, FL, USA

 

[berkshire]

kawaminaii
I am having a problem relating to the size of your curved beam " The radius of curvature of the resultant beam should be 2.5 meters and its span will be 8 meters", a 180 degree arc would only span 5 Metres. Do you mean an 8 metre span with a rise of 2.5 metres ? That would be a 9.95M curved arc over 128 degrees spanning 8M. having a radius of 4.45 M.
B.E.

You are judged not by what you know, but by what you can do.

 

[KawaMinaii]

ِDear berkshire, you are right. The 2.5 m of radius is not correct. The radius of the curvature should be 5.34 m which its height would be 1.8 m.
Thank you for the notification.

 

[berkshire]

Ok,
Now you are describing an arc which should be doable by cold forming in 6061-T6.
B.E.

You are judged not by what you know, but by what you can do.

 

[ruhlich]

Thanks to compositepro & Berkshire for your comments regarding my 6061 forming practices. I might have known some of this 50 years ago when I studied Heat Transfer, but I'm afraid it's since left me.